| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 13750, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0036363636363636364, | |
| "grad_norm": 8.730692897709844, | |
| "learning_rate": 1.0827317615529767e-06, | |
| "loss": 0.1943, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.007272727272727273, | |
| "grad_norm": 6.563599993841062, | |
| "learning_rate": 1.2745743112470354e-06, | |
| "loss": 0.1472, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01090909090909091, | |
| "grad_norm": 11.109271986600454, | |
| "learning_rate": 1.386795008860794e-06, | |
| "loss": 0.1713, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.014545454545454545, | |
| "grad_norm": 5.772674964443883, | |
| "learning_rate": 1.4664168609410934e-06, | |
| "loss": 0.1585, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.01818181818181818, | |
| "grad_norm": 14.76596653457522, | |
| "learning_rate": 1.5281763674824356e-06, | |
| "loss": 0.1429, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.02181818181818182, | |
| "grad_norm": 9.568061757019173, | |
| "learning_rate": 1.5786375585548526e-06, | |
| "loss": 0.1316, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.025454545454545455, | |
| "grad_norm": 10.301645195685637, | |
| "learning_rate": 1.621301887696672e-06, | |
| "loss": 0.1305, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.02909090909090909, | |
| "grad_norm": 5.6993526158027645, | |
| "learning_rate": 1.6582594106351517e-06, | |
| "loss": 0.109, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.03272727272727273, | |
| "grad_norm": 4.915542646471226, | |
| "learning_rate": 1.6908582561686116e-06, | |
| "loss": 0.1269, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.03636363636363636, | |
| "grad_norm": 9.392642904031657, | |
| "learning_rate": 1.720018917176494e-06, | |
| "loss": 0.1201, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 5.047170718950901, | |
| "learning_rate": 1.746397943764599e-06, | |
| "loss": 0.1304, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.04363636363636364, | |
| "grad_norm": 6.865869458534065, | |
| "learning_rate": 1.7704801082489109e-06, | |
| "loss": 0.1185, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.04727272727272727, | |
| "grad_norm": 7.641476199550763, | |
| "learning_rate": 1.7926335520703266e-06, | |
| "loss": 0.1071, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.05090909090909091, | |
| "grad_norm": 5.401679074654614, | |
| "learning_rate": 1.8131444373907304e-06, | |
| "loss": 0.1153, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.05454545454545454, | |
| "grad_norm": 6.576489460422398, | |
| "learning_rate": 1.8322396147902533e-06, | |
| "loss": 0.1104, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.05818181818181818, | |
| "grad_norm": 7.069491006747404, | |
| "learning_rate": 1.8501019603292104e-06, | |
| "loss": 0.1164, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.06181818181818182, | |
| "grad_norm": 5.9290634450105975, | |
| "learning_rate": 1.866881054798162e-06, | |
| "loss": 0.1219, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.06545454545454546, | |
| "grad_norm": 5.426281987614212, | |
| "learning_rate": 1.88270080586267e-06, | |
| "loss": 0.1215, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.06909090909090909, | |
| "grad_norm": 7.8695827198040105, | |
| "learning_rate": 1.8976650066475357e-06, | |
| "loss": 0.1016, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.07272727272727272, | |
| "grad_norm": 16.511242359906724, | |
| "learning_rate": 1.9118614668705526e-06, | |
| "loss": 0.1005, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.07636363636363637, | |
| "grad_norm": 5.456737119903023, | |
| "learning_rate": 1.9253651350044893e-06, | |
| "loss": 0.0967, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.0750937470823105, | |
| "learning_rate": 1.9382404934586575e-06, | |
| "loss": 0.0989, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.08363636363636363, | |
| "grad_norm": 8.699529648266866, | |
| "learning_rate": 1.9505434209019962e-06, | |
| "loss": 0.0988, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.08727272727272728, | |
| "grad_norm": 6.77391900591751, | |
| "learning_rate": 1.9623226579429693e-06, | |
| "loss": 0.1078, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 6.290303703702585, | |
| "learning_rate": 1.973620973411895e-06, | |
| "loss": 0.1118, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.09454545454545454, | |
| "grad_norm": 4.7489952229331, | |
| "learning_rate": 1.984476101764385e-06, | |
| "loss": 0.1014, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.09818181818181818, | |
| "grad_norm": 5.92022086118128, | |
| "learning_rate": 1.9949215034764293e-06, | |
| "loss": 0.0865, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.10181818181818182, | |
| "grad_norm": 3.9298710865494013, | |
| "learning_rate": 1.9961212121212122e-06, | |
| "loss": 0.1019, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.10545454545454545, | |
| "grad_norm": 1.35952985911884, | |
| "learning_rate": 1.9880404040404038e-06, | |
| "loss": 0.0889, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.10909090909090909, | |
| "grad_norm": 6.074760388330947, | |
| "learning_rate": 1.9799595959595958e-06, | |
| "loss": 0.088, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.11272727272727273, | |
| "grad_norm": 3.281001635643186, | |
| "learning_rate": 1.9718787878787877e-06, | |
| "loss": 0.0939, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.11636363636363636, | |
| "grad_norm": 2.834658163409011, | |
| "learning_rate": 1.9637979797979797e-06, | |
| "loss": 0.0941, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 5.716408171893896, | |
| "learning_rate": 1.9557171717171717e-06, | |
| "loss": 0.0869, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.12363636363636364, | |
| "grad_norm": 4.355380070215489, | |
| "learning_rate": 1.9476363636363637e-06, | |
| "loss": 0.0879, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.12727272727272726, | |
| "grad_norm": 2.56506898084542, | |
| "learning_rate": 1.939555555555555e-06, | |
| "loss": 0.0833, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.13090909090909092, | |
| "grad_norm": 3.1998963994898744, | |
| "learning_rate": 1.9314747474747476e-06, | |
| "loss": 0.0827, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.13454545454545455, | |
| "grad_norm": 2.993946484706046, | |
| "learning_rate": 1.923393939393939e-06, | |
| "loss": 0.0884, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.13818181818181818, | |
| "grad_norm": 4.066355592263713, | |
| "learning_rate": 1.915313131313131e-06, | |
| "loss": 0.0843, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.14181818181818182, | |
| "grad_norm": 3.9444301109471422, | |
| "learning_rate": 1.9072323232323231e-06, | |
| "loss": 0.079, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.14545454545454545, | |
| "grad_norm": 4.998815586950459, | |
| "learning_rate": 1.899151515151515e-06, | |
| "loss": 0.0742, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.14909090909090908, | |
| "grad_norm": 4.98600431497313, | |
| "learning_rate": 1.891070707070707e-06, | |
| "loss": 0.0883, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.15272727272727274, | |
| "grad_norm": 7.068049048520739, | |
| "learning_rate": 1.8829898989898988e-06, | |
| "loss": 0.0888, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.15636363636363637, | |
| "grad_norm": 6.445827098301751, | |
| "learning_rate": 1.8749090909090908e-06, | |
| "loss": 0.083, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.422103821997445, | |
| "learning_rate": 1.8668282828282826e-06, | |
| "loss": 0.0784, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.16363636363636364, | |
| "grad_norm": 10.593030877025758, | |
| "learning_rate": 1.8589090909090909e-06, | |
| "loss": 0.0907, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.16727272727272727, | |
| "grad_norm": 5.73637966279623, | |
| "learning_rate": 1.8508282828282828e-06, | |
| "loss": 0.0745, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.1709090909090909, | |
| "grad_norm": 4.435065232821809, | |
| "learning_rate": 1.8427474747474748e-06, | |
| "loss": 0.0764, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.17454545454545456, | |
| "grad_norm": 6.932446538015379, | |
| "learning_rate": 1.8346666666666666e-06, | |
| "loss": 0.082, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.1781818181818182, | |
| "grad_norm": 8.365214538194435, | |
| "learning_rate": 1.8265858585858586e-06, | |
| "loss": 0.0792, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 3.8114318602084714, | |
| "learning_rate": 1.8185050505050503e-06, | |
| "loss": 0.0819, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.18545454545454546, | |
| "grad_norm": 8.151155598162092, | |
| "learning_rate": 1.8104242424242425e-06, | |
| "loss": 0.0796, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.1890909090909091, | |
| "grad_norm": 3.853054291632494, | |
| "learning_rate": 1.8023434343434343e-06, | |
| "loss": 0.0724, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.19272727272727272, | |
| "grad_norm": 2.7845888134928454, | |
| "learning_rate": 1.7942626262626262e-06, | |
| "loss": 0.0778, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.19636363636363635, | |
| "grad_norm": 6.178204576735234, | |
| "learning_rate": 1.786181818181818e-06, | |
| "loss": 0.0852, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 3.2752914823154926, | |
| "learning_rate": 1.7781010101010102e-06, | |
| "loss": 0.0724, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.20363636363636364, | |
| "grad_norm": 4.209509329574692, | |
| "learning_rate": 1.770020202020202e-06, | |
| "loss": 0.0793, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.20727272727272728, | |
| "grad_norm": 6.139612323065608, | |
| "learning_rate": 1.761939393939394e-06, | |
| "loss": 0.07, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.2109090909090909, | |
| "grad_norm": 2.455643448424175, | |
| "learning_rate": 1.7538585858585857e-06, | |
| "loss": 0.0738, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.21454545454545454, | |
| "grad_norm": 5.4333976968022695, | |
| "learning_rate": 1.7457777777777779e-06, | |
| "loss": 0.0783, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.21818181818181817, | |
| "grad_norm": 5.0280981201819, | |
| "learning_rate": 1.7376969696969696e-06, | |
| "loss": 0.0615, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.22181818181818183, | |
| "grad_norm": 6.437028886166945, | |
| "learning_rate": 1.7296161616161616e-06, | |
| "loss": 0.0766, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.22545454545454546, | |
| "grad_norm": 2.5493783641885575, | |
| "learning_rate": 1.7215353535353534e-06, | |
| "loss": 0.0667, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.2290909090909091, | |
| "grad_norm": 7.176307787653395, | |
| "learning_rate": 1.7134545454545456e-06, | |
| "loss": 0.0814, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.23272727272727273, | |
| "grad_norm": 6.4838446342779585, | |
| "learning_rate": 1.7053737373737373e-06, | |
| "loss": 0.0804, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.23636363636363636, | |
| "grad_norm": 6.05439389409086, | |
| "learning_rate": 1.6972929292929293e-06, | |
| "loss": 0.0758, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 4.8914160530492925, | |
| "learning_rate": 1.6893737373737374e-06, | |
| "loss": 0.0726, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.24363636363636362, | |
| "grad_norm": 5.1032181254180875, | |
| "learning_rate": 1.6812929292929292e-06, | |
| "loss": 0.0742, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.24727272727272728, | |
| "grad_norm": 5.512044376565606, | |
| "learning_rate": 1.6732121212121211e-06, | |
| "loss": 0.0721, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.2509090909090909, | |
| "grad_norm": 6.215671409867116, | |
| "learning_rate": 1.6651313131313131e-06, | |
| "loss": 0.0714, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.2545454545454545, | |
| "grad_norm": 4.934931799389331, | |
| "learning_rate": 1.657050505050505e-06, | |
| "loss": 0.0834, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.2581818181818182, | |
| "grad_norm": 5.766772809824982, | |
| "learning_rate": 1.6489696969696969e-06, | |
| "loss": 0.0719, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.26181818181818184, | |
| "grad_norm": 2.3502736155525907, | |
| "learning_rate": 1.6408888888888888e-06, | |
| "loss": 0.0667, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.26545454545454544, | |
| "grad_norm": 5.212095666658788, | |
| "learning_rate": 1.6328080808080806e-06, | |
| "loss": 0.0657, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.2690909090909091, | |
| "grad_norm": 2.2102868298776994, | |
| "learning_rate": 1.6247272727272728e-06, | |
| "loss": 0.0746, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 7.107367934409905, | |
| "learning_rate": 1.6166464646464645e-06, | |
| "loss": 0.0686, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.27636363636363637, | |
| "grad_norm": 4.9640350043256065, | |
| "learning_rate": 1.6085656565656565e-06, | |
| "loss": 0.0715, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 3.8508170066035556, | |
| "learning_rate": 1.6004848484848483e-06, | |
| "loss": 0.0755, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.28363636363636363, | |
| "grad_norm": 6.406204149035612, | |
| "learning_rate": 1.5924040404040405e-06, | |
| "loss": 0.0722, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.2872727272727273, | |
| "grad_norm": 6.4755620057100005, | |
| "learning_rate": 1.5843232323232322e-06, | |
| "loss": 0.0653, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 4.7041545767525985, | |
| "learning_rate": 1.5762424242424242e-06, | |
| "loss": 0.061, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.29454545454545455, | |
| "grad_norm": 6.597148868978393, | |
| "learning_rate": 1.568161616161616e-06, | |
| "loss": 0.0631, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.29818181818181816, | |
| "grad_norm": 4.24188377459276, | |
| "learning_rate": 1.5600808080808082e-06, | |
| "loss": 0.0767, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.3018181818181818, | |
| "grad_norm": 5.185458229586008, | |
| "learning_rate": 1.552e-06, | |
| "loss": 0.0768, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.3054545454545455, | |
| "grad_norm": 7.282950927099741, | |
| "learning_rate": 1.543919191919192e-06, | |
| "loss": 0.0696, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.3090909090909091, | |
| "grad_norm": 6.602554878720333, | |
| "learning_rate": 1.5358383838383837e-06, | |
| "loss": 0.0644, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.31272727272727274, | |
| "grad_norm": 5.8760748346703755, | |
| "learning_rate": 1.5277575757575756e-06, | |
| "loss": 0.0629, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.31636363636363635, | |
| "grad_norm": 5.87511103099038, | |
| "learning_rate": 1.5198383838383837e-06, | |
| "loss": 0.0699, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 5.07793513016435, | |
| "learning_rate": 1.5117575757575757e-06, | |
| "loss": 0.0744, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.3236363636363636, | |
| "grad_norm": 3.3381013307087595, | |
| "learning_rate": 1.5036767676767677e-06, | |
| "loss": 0.0671, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.32727272727272727, | |
| "grad_norm": 6.087362688149139, | |
| "learning_rate": 1.4955959595959594e-06, | |
| "loss": 0.0687, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.33090909090909093, | |
| "grad_norm": 3.459475929834813, | |
| "learning_rate": 1.4875151515151514e-06, | |
| "loss": 0.0685, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.33454545454545453, | |
| "grad_norm": 4.891090834007153, | |
| "learning_rate": 1.4794343434343432e-06, | |
| "loss": 0.0725, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.3381818181818182, | |
| "grad_norm": 4.691041549495357, | |
| "learning_rate": 1.4713535353535354e-06, | |
| "loss": 0.061, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.3418181818181818, | |
| "grad_norm": 5.924562513722335, | |
| "learning_rate": 1.4632727272727271e-06, | |
| "loss": 0.0758, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.34545454545454546, | |
| "grad_norm": 4.580382327650832, | |
| "learning_rate": 1.4551919191919191e-06, | |
| "loss": 0.0743, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.3490909090909091, | |
| "grad_norm": 4.6681099203941026, | |
| "learning_rate": 1.4471111111111109e-06, | |
| "loss": 0.0656, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.3527272727272727, | |
| "grad_norm": 5.030610606632313, | |
| "learning_rate": 1.439030303030303e-06, | |
| "loss": 0.0737, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.3563636363636364, | |
| "grad_norm": 7.614614950202259, | |
| "learning_rate": 1.4309494949494948e-06, | |
| "loss": 0.0686, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 5.981464365403329, | |
| "learning_rate": 1.4228686868686868e-06, | |
| "loss": 0.0748, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 4.277903564607442, | |
| "learning_rate": 1.4147878787878786e-06, | |
| "loss": 0.0568, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.36727272727272725, | |
| "grad_norm": 17.93564630692927, | |
| "learning_rate": 1.4067070707070707e-06, | |
| "loss": 0.0663, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.3709090909090909, | |
| "grad_norm": 2.2042619733871374, | |
| "learning_rate": 1.3986262626262625e-06, | |
| "loss": 0.0648, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.37454545454545457, | |
| "grad_norm": 4.712133353867323, | |
| "learning_rate": 1.3905454545454545e-06, | |
| "loss": 0.0773, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.3781818181818182, | |
| "grad_norm": 5.573958863346438, | |
| "learning_rate": 1.3824646464646462e-06, | |
| "loss": 0.0679, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.38181818181818183, | |
| "grad_norm": 4.848377798083522, | |
| "learning_rate": 1.3743838383838384e-06, | |
| "loss": 0.0671, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.38545454545454544, | |
| "grad_norm": 5.980348193388087, | |
| "learning_rate": 1.3663030303030302e-06, | |
| "loss": 0.0694, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.3890909090909091, | |
| "grad_norm": 2.672380009268972, | |
| "learning_rate": 1.3582222222222222e-06, | |
| "loss": 0.0676, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.3927272727272727, | |
| "grad_norm": 2.964040453903747, | |
| "learning_rate": 1.350141414141414e-06, | |
| "loss": 0.0643, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.39636363636363636, | |
| "grad_norm": 7.038083972721559, | |
| "learning_rate": 1.342060606060606e-06, | |
| "loss": 0.0632, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 3.8224124558617496, | |
| "learning_rate": 1.3339797979797979e-06, | |
| "loss": 0.0731, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.4036363636363636, | |
| "grad_norm": 7.048096138652444, | |
| "learning_rate": 1.3258989898989899e-06, | |
| "loss": 0.0594, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.4072727272727273, | |
| "grad_norm": 5.745560964316516, | |
| "learning_rate": 1.3178181818181816e-06, | |
| "loss": 0.0532, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.4109090909090909, | |
| "grad_norm": 4.188291560272238, | |
| "learning_rate": 1.3097373737373736e-06, | |
| "loss": 0.0633, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.41454545454545455, | |
| "grad_norm": 7.49562784549071, | |
| "learning_rate": 1.3018181818181817e-06, | |
| "loss": 0.0681, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.41818181818181815, | |
| "grad_norm": 3.5275023161774404, | |
| "learning_rate": 1.2937373737373737e-06, | |
| "loss": 0.0664, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.4218181818181818, | |
| "grad_norm": 4.119155442422377, | |
| "learning_rate": 1.2856565656565656e-06, | |
| "loss": 0.0617, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.4254545454545455, | |
| "grad_norm": 3.065330909676055, | |
| "learning_rate": 1.2775757575757576e-06, | |
| "loss": 0.055, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.4290909090909091, | |
| "grad_norm": 5.499516032340102, | |
| "learning_rate": 1.2694949494949494e-06, | |
| "loss": 0.0666, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.43272727272727274, | |
| "grad_norm": 7.319764009851145, | |
| "learning_rate": 1.2614141414141414e-06, | |
| "loss": 0.0679, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.43636363636363634, | |
| "grad_norm": 3.842328064446604, | |
| "learning_rate": 1.2533333333333333e-06, | |
| "loss": 0.0627, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.6429715042707909, | |
| "learning_rate": 1.2452525252525253e-06, | |
| "loss": 0.0566, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.44363636363636366, | |
| "grad_norm": 3.3628676599920797, | |
| "learning_rate": 1.237171717171717e-06, | |
| "loss": 0.0678, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.44727272727272727, | |
| "grad_norm": 2.2255401941396613, | |
| "learning_rate": 1.229090909090909e-06, | |
| "loss": 0.0536, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.4509090909090909, | |
| "grad_norm": 4.667731753921009, | |
| "learning_rate": 1.221010101010101e-06, | |
| "loss": 0.0704, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 4.3551424870793465, | |
| "learning_rate": 1.212929292929293e-06, | |
| "loss": 0.0637, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.4581818181818182, | |
| "grad_norm": 3.3583233273895865, | |
| "learning_rate": 1.2048484848484848e-06, | |
| "loss": 0.0627, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.4618181818181818, | |
| "grad_norm": 4.2019753851553165, | |
| "learning_rate": 1.1967676767676767e-06, | |
| "loss": 0.0549, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.46545454545454545, | |
| "grad_norm": 2.5872795563931024, | |
| "learning_rate": 1.1886868686868687e-06, | |
| "loss": 0.0544, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.4690909090909091, | |
| "grad_norm": 7.041225020724765, | |
| "learning_rate": 1.1806060606060607e-06, | |
| "loss": 0.06, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.4727272727272727, | |
| "grad_norm": 5.958896992660691, | |
| "learning_rate": 1.1725252525252524e-06, | |
| "loss": 0.0564, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.4763636363636364, | |
| "grad_norm": 4.476780431745298, | |
| "learning_rate": 1.1644444444444444e-06, | |
| "loss": 0.053, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.356776109954863, | |
| "learning_rate": 1.1565252525252525e-06, | |
| "loss": 0.0601, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.48363636363636364, | |
| "grad_norm": 8.192732907739552, | |
| "learning_rate": 1.1484444444444443e-06, | |
| "loss": 0.0569, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.48727272727272725, | |
| "grad_norm": 4.683769267327389, | |
| "learning_rate": 1.1403636363636365e-06, | |
| "loss": 0.0573, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.4909090909090909, | |
| "grad_norm": 2.2927166931841096, | |
| "learning_rate": 1.1322828282828282e-06, | |
| "loss": 0.0555, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.49454545454545457, | |
| "grad_norm": 5.921072631266765, | |
| "learning_rate": 1.1242020202020202e-06, | |
| "loss": 0.0751, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.49818181818181817, | |
| "grad_norm": 6.639001409977233, | |
| "learning_rate": 1.116121212121212e-06, | |
| "loss": 0.0585, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.5018181818181818, | |
| "grad_norm": 4.5171318312031525, | |
| "learning_rate": 1.108040404040404e-06, | |
| "loss": 0.0598, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.5054545454545455, | |
| "grad_norm": 3.889837713901535, | |
| "learning_rate": 1.099959595959596e-06, | |
| "loss": 0.0557, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.509090909090909, | |
| "grad_norm": 4.7778306090713905, | |
| "learning_rate": 1.091878787878788e-06, | |
| "loss": 0.069, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.5127272727272727, | |
| "grad_norm": 7.318064341364163, | |
| "learning_rate": 1.0837979797979797e-06, | |
| "loss": 0.0589, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.5163636363636364, | |
| "grad_norm": 0.8134545807541753, | |
| "learning_rate": 1.0757171717171716e-06, | |
| "loss": 0.0568, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 8.47016607740995, | |
| "learning_rate": 1.0676363636363636e-06, | |
| "loss": 0.065, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.5236363636363637, | |
| "grad_norm": 4.216730157024882, | |
| "learning_rate": 1.0595555555555556e-06, | |
| "loss": 0.0551, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.5272727272727272, | |
| "grad_norm": 8.409317988298476, | |
| "learning_rate": 1.0514747474747473e-06, | |
| "loss": 0.0591, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 0.5309090909090909, | |
| "grad_norm": 3.8221711861975938, | |
| "learning_rate": 1.0433939393939393e-06, | |
| "loss": 0.0652, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.5345454545454545, | |
| "grad_norm": 5.941432574111638, | |
| "learning_rate": 1.0353131313131313e-06, | |
| "loss": 0.0555, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 0.5381818181818182, | |
| "grad_norm": 3.7955500798877027, | |
| "learning_rate": 1.0272323232323233e-06, | |
| "loss": 0.0639, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.5418181818181819, | |
| "grad_norm": 5.077036391729846, | |
| "learning_rate": 1.019151515151515e-06, | |
| "loss": 0.0574, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 2.967208140674572, | |
| "learning_rate": 1.011070707070707e-06, | |
| "loss": 0.055, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.5490909090909091, | |
| "grad_norm": 6.4064840134469065, | |
| "learning_rate": 1.002989898989899e-06, | |
| "loss": 0.0563, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 0.5527272727272727, | |
| "grad_norm": 8.224100322274762, | |
| "learning_rate": 9.949090909090908e-07, | |
| "loss": 0.0542, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.5563636363636364, | |
| "grad_norm": 7.195740329809225, | |
| "learning_rate": 9.868282828282827e-07, | |
| "loss": 0.0518, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.4648844395390421, | |
| "learning_rate": 9.787474747474747e-07, | |
| "loss": 0.0513, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.5636363636363636, | |
| "grad_norm": 1.2171654793506612, | |
| "learning_rate": 9.706666666666667e-07, | |
| "loss": 0.0611, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 0.5672727272727273, | |
| "grad_norm": 3.5164849479935687, | |
| "learning_rate": 9.625858585858584e-07, | |
| "loss": 0.0637, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.5709090909090909, | |
| "grad_norm": 7.553161151311784, | |
| "learning_rate": 9.545050505050504e-07, | |
| "loss": 0.0566, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 0.5745454545454546, | |
| "grad_norm": 5.745398629139221, | |
| "learning_rate": 9.464242424242424e-07, | |
| "loss": 0.065, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.5781818181818181, | |
| "grad_norm": 3.814619257014882, | |
| "learning_rate": 9.383434343434344e-07, | |
| "loss": 0.0598, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 0.5818181818181818, | |
| "grad_norm": 4.131828897750148, | |
| "learning_rate": 9.302626262626262e-07, | |
| "loss": 0.0712, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.5854545454545454, | |
| "grad_norm": 3.4914242723281075, | |
| "learning_rate": 9.221818181818182e-07, | |
| "loss": 0.0608, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 0.5890909090909091, | |
| "grad_norm": 5.282146490628778, | |
| "learning_rate": 9.141010101010101e-07, | |
| "loss": 0.0555, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.5927272727272728, | |
| "grad_norm": 6.151329591511757, | |
| "learning_rate": 9.060202020202021e-07, | |
| "loss": 0.0597, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 0.5963636363636363, | |
| "grad_norm": 0.5491620334300515, | |
| "learning_rate": 8.979393939393939e-07, | |
| "loss": 0.0525, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 6.024173360125982, | |
| "learning_rate": 8.898585858585859e-07, | |
| "loss": 0.0622, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 0.6036363636363636, | |
| "grad_norm": 7.030861805690174, | |
| "learning_rate": 8.817777777777778e-07, | |
| "loss": 0.0556, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.6072727272727273, | |
| "grad_norm": 5.380176092336188, | |
| "learning_rate": 8.736969696969697e-07, | |
| "loss": 0.0542, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 0.610909090909091, | |
| "grad_norm": 5.417266910258118, | |
| "learning_rate": 8.656161616161616e-07, | |
| "loss": 0.0492, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.6145454545454545, | |
| "grad_norm": 4.935151132976808, | |
| "learning_rate": 8.575353535353535e-07, | |
| "loss": 0.0492, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 0.6181818181818182, | |
| "grad_norm": 4.659946484892871, | |
| "learning_rate": 8.494545454545455e-07, | |
| "loss": 0.0571, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.6218181818181818, | |
| "grad_norm": 6.771681262010034, | |
| "learning_rate": 8.413737373737373e-07, | |
| "loss": 0.0599, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 0.6254545454545455, | |
| "grad_norm": 6.08702835668505, | |
| "learning_rate": 8.332929292929293e-07, | |
| "loss": 0.0685, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.6290909090909091, | |
| "grad_norm": 6.534364094809468, | |
| "learning_rate": 8.252121212121212e-07, | |
| "loss": 0.0546, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 0.6327272727272727, | |
| "grad_norm": 3.834673453780823, | |
| "learning_rate": 8.171313131313131e-07, | |
| "loss": 0.0597, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 6.698669517846524, | |
| "learning_rate": 8.09050505050505e-07, | |
| "loss": 0.056, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 5.4043455793793775, | |
| "learning_rate": 8.00969696969697e-07, | |
| "loss": 0.0613, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.6436363636363637, | |
| "grad_norm": 1.660910042565409, | |
| "learning_rate": 7.928888888888889e-07, | |
| "loss": 0.0598, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 0.6472727272727272, | |
| "grad_norm": 7.795287679730981, | |
| "learning_rate": 7.848080808080808e-07, | |
| "loss": 0.0507, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.6509090909090909, | |
| "grad_norm": 3.293884222491616, | |
| "learning_rate": 7.767272727272727e-07, | |
| "loss": 0.0505, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 0.6545454545454545, | |
| "grad_norm": 4.230324246940526, | |
| "learning_rate": 7.686464646464647e-07, | |
| "loss": 0.048, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.6581818181818182, | |
| "grad_norm": 3.140145943767063, | |
| "learning_rate": 7.605656565656565e-07, | |
| "loss": 0.0641, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 0.6618181818181819, | |
| "grad_norm": 4.863206340412036, | |
| "learning_rate": 7.526464646464646e-07, | |
| "loss": 0.0598, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.6654545454545454, | |
| "grad_norm": 3.458397202112644, | |
| "learning_rate": 7.445656565656565e-07, | |
| "loss": 0.0483, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 0.6690909090909091, | |
| "grad_norm": 4.78959754830013, | |
| "learning_rate": 7.364848484848485e-07, | |
| "loss": 0.0536, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.6727272727272727, | |
| "grad_norm": 10.740436647088291, | |
| "learning_rate": 7.284040404040404e-07, | |
| "loss": 0.0572, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 0.6763636363636364, | |
| "grad_norm": 1.6489558557462445, | |
| "learning_rate": 7.204848484848484e-07, | |
| "loss": 0.0604, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 7.3316901995966335, | |
| "learning_rate": 7.124040404040403e-07, | |
| "loss": 0.0559, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 0.6836363636363636, | |
| "grad_norm": 1.5208003767788665, | |
| "learning_rate": 7.043232323232323e-07, | |
| "loss": 0.0523, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.6872727272727273, | |
| "grad_norm": 3.7519776533636673, | |
| "learning_rate": 6.962424242424242e-07, | |
| "loss": 0.0482, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 0.6909090909090909, | |
| "grad_norm": 1.6222324278842861, | |
| "learning_rate": 6.881616161616161e-07, | |
| "loss": 0.0572, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.6945454545454546, | |
| "grad_norm": 7.0411497148292925, | |
| "learning_rate": 6.80080808080808e-07, | |
| "loss": 0.0571, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 0.6981818181818182, | |
| "grad_norm": 2.6841788063694616, | |
| "learning_rate": 6.72e-07, | |
| "loss": 0.0564, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.7018181818181818, | |
| "grad_norm": 6.819000782187993, | |
| "learning_rate": 6.639191919191919e-07, | |
| "loss": 0.0605, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 0.7054545454545454, | |
| "grad_norm": 7.92210806178369, | |
| "learning_rate": 6.558383838383838e-07, | |
| "loss": 0.0548, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.7090909090909091, | |
| "grad_norm": 4.811455749568285, | |
| "learning_rate": 6.477575757575757e-07, | |
| "loss": 0.0528, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 0.7127272727272728, | |
| "grad_norm": 5.308321038455987, | |
| "learning_rate": 6.396767676767676e-07, | |
| "loss": 0.0527, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.7163636363636363, | |
| "grad_norm": 5.069220520833826, | |
| "learning_rate": 6.315959595959595e-07, | |
| "loss": 0.0566, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 5.151662369356777, | |
| "learning_rate": 6.235151515151514e-07, | |
| "loss": 0.0567, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.7236363636363636, | |
| "grad_norm": 6.063613002898996, | |
| "learning_rate": 6.154343434343434e-07, | |
| "loss": 0.0587, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.0628386687029727, | |
| "learning_rate": 6.073535353535353e-07, | |
| "loss": 0.0539, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.730909090909091, | |
| "grad_norm": 4.641899576056589, | |
| "learning_rate": 5.992727272727272e-07, | |
| "loss": 0.0516, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 0.7345454545454545, | |
| "grad_norm": 4.650551400255244, | |
| "learning_rate": 5.911919191919191e-07, | |
| "loss": 0.0471, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 0.7381818181818182, | |
| "grad_norm": 4.567296054840706, | |
| "learning_rate": 5.831111111111111e-07, | |
| "loss": 0.0479, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 0.7418181818181818, | |
| "grad_norm": 5.044836996902614, | |
| "learning_rate": 5.750303030303029e-07, | |
| "loss": 0.0561, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 0.7454545454545455, | |
| "grad_norm": 4.248950379851986, | |
| "learning_rate": 5.669494949494949e-07, | |
| "loss": 0.0601, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 0.7490909090909091, | |
| "grad_norm": 2.028141754034167, | |
| "learning_rate": 5.588686868686868e-07, | |
| "loss": 0.0422, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 0.7527272727272727, | |
| "grad_norm": 10.007581888321717, | |
| "learning_rate": 5.507878787878788e-07, | |
| "loss": 0.0469, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 0.7563636363636363, | |
| "grad_norm": 4.577989595773856, | |
| "learning_rate": 5.427070707070706e-07, | |
| "loss": 0.05, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 6.521133984672062, | |
| "learning_rate": 5.346262626262626e-07, | |
| "loss": 0.0428, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 0.7636363636363637, | |
| "grad_norm": 3.9509494509580705, | |
| "learning_rate": 5.265454545454545e-07, | |
| "loss": 0.0522, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.7672727272727272, | |
| "grad_norm": 6.944726977217027, | |
| "learning_rate": 5.184646464646465e-07, | |
| "loss": 0.0515, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 0.7709090909090909, | |
| "grad_norm": 3.831893622894902, | |
| "learning_rate": 5.103838383838383e-07, | |
| "loss": 0.0541, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 0.7745454545454545, | |
| "grad_norm": 4.8538480771451695, | |
| "learning_rate": 5.023030303030302e-07, | |
| "loss": 0.06, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 0.7781818181818182, | |
| "grad_norm": 9.741124343678399, | |
| "learning_rate": 4.942222222222222e-07, | |
| "loss": 0.0522, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 0.7818181818181819, | |
| "grad_norm": 1.637375802142195, | |
| "learning_rate": 4.861414141414141e-07, | |
| "loss": 0.0479, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 0.7854545454545454, | |
| "grad_norm": 2.3084795257142936, | |
| "learning_rate": 4.78060606060606e-07, | |
| "loss": 0.0626, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 0.7890909090909091, | |
| "grad_norm": 6.674598298862028, | |
| "learning_rate": 4.69979797979798e-07, | |
| "loss": 0.0476, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 0.7927272727272727, | |
| "grad_norm": 5.520833007044753, | |
| "learning_rate": 4.618989898989899e-07, | |
| "loss": 0.0473, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 0.7963636363636364, | |
| "grad_norm": 7.535471892284138, | |
| "learning_rate": 4.5381818181818183e-07, | |
| "loss": 0.0541, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.119845815305954, | |
| "learning_rate": 4.457373737373737e-07, | |
| "loss": 0.0469, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.8036363636363636, | |
| "grad_norm": 4.865422744810339, | |
| "learning_rate": 4.376565656565656e-07, | |
| "loss": 0.0542, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 0.8072727272727273, | |
| "grad_norm": 4.381485410017174, | |
| "learning_rate": 4.2957575757575754e-07, | |
| "loss": 0.0595, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 0.8109090909090909, | |
| "grad_norm": 4.763162545803381, | |
| "learning_rate": 4.2149494949494947e-07, | |
| "loss": 0.0542, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 0.8145454545454546, | |
| "grad_norm": 3.986784442089026, | |
| "learning_rate": 4.134141414141414e-07, | |
| "loss": 0.0533, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 1.9368964454962925, | |
| "learning_rate": 4.053333333333333e-07, | |
| "loss": 0.051, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 0.8218181818181818, | |
| "grad_norm": 5.557279558773371, | |
| "learning_rate": 3.9725252525252523e-07, | |
| "loss": 0.0507, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 0.8254545454545454, | |
| "grad_norm": 5.301180819559454, | |
| "learning_rate": 3.8917171717171715e-07, | |
| "loss": 0.0478, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 0.8290909090909091, | |
| "grad_norm": 6.393245840471723, | |
| "learning_rate": 3.810909090909091e-07, | |
| "loss": 0.0541, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 0.8327272727272728, | |
| "grad_norm": 5.140132428255607, | |
| "learning_rate": 3.731717171717171e-07, | |
| "loss": 0.0574, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 0.8363636363636363, | |
| "grad_norm": 4.020573635641951, | |
| "learning_rate": 3.6509090909090904e-07, | |
| "loss": 0.0488, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 8.669170555909107, | |
| "learning_rate": 3.5701010101010096e-07, | |
| "loss": 0.0459, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 0.8436363636363636, | |
| "grad_norm": 2.7227046748798425, | |
| "learning_rate": 3.489292929292929e-07, | |
| "loss": 0.0427, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 0.8472727272727273, | |
| "grad_norm": 8.264092633240582, | |
| "learning_rate": 3.408484848484848e-07, | |
| "loss": 0.0443, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 0.850909090909091, | |
| "grad_norm": 2.2161705174100543, | |
| "learning_rate": 3.3276767676767673e-07, | |
| "loss": 0.05, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 0.8545454545454545, | |
| "grad_norm": 4.049181895573167, | |
| "learning_rate": 3.2468686868686865e-07, | |
| "loss": 0.045, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 0.8581818181818182, | |
| "grad_norm": 6.448219337405561, | |
| "learning_rate": 3.1660606060606057e-07, | |
| "loss": 0.0445, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 0.8618181818181818, | |
| "grad_norm": 1.6198743531280058, | |
| "learning_rate": 3.0852525252525255e-07, | |
| "loss": 0.0436, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 0.8654545454545455, | |
| "grad_norm": 5.887817361300306, | |
| "learning_rate": 3.0044444444444447e-07, | |
| "loss": 0.058, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 0.8690909090909091, | |
| "grad_norm": 5.420521679930149, | |
| "learning_rate": 2.923636363636364e-07, | |
| "loss": 0.0578, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 0.8727272727272727, | |
| "grad_norm": 6.9211836011036105, | |
| "learning_rate": 2.842828282828283e-07, | |
| "loss": 0.0468, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.8763636363636363, | |
| "grad_norm": 6.962499989042162, | |
| "learning_rate": 2.7620202020202024e-07, | |
| "loss": 0.0565, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 2.251073369325601, | |
| "learning_rate": 2.681212121212121e-07, | |
| "loss": 0.051, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 0.8836363636363637, | |
| "grad_norm": 6.051100878218682, | |
| "learning_rate": 2.6004040404040403e-07, | |
| "loss": 0.0494, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 0.8872727272727273, | |
| "grad_norm": 5.559109629686928, | |
| "learning_rate": 2.5195959595959595e-07, | |
| "loss": 0.0508, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 0.8909090909090909, | |
| "grad_norm": 2.7004728150520614, | |
| "learning_rate": 2.4387878787878787e-07, | |
| "loss": 0.0514, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 0.8945454545454545, | |
| "grad_norm": 7.949939262110135, | |
| "learning_rate": 2.3579797979797977e-07, | |
| "loss": 0.0573, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 0.8981818181818182, | |
| "grad_norm": 6.02491008010814, | |
| "learning_rate": 2.277171717171717e-07, | |
| "loss": 0.0526, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 0.9018181818181819, | |
| "grad_norm": 2.893965651713638, | |
| "learning_rate": 2.196363636363636e-07, | |
| "loss": 0.0553, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 0.9054545454545454, | |
| "grad_norm": 0.7646769278371894, | |
| "learning_rate": 2.1155555555555554e-07, | |
| "loss": 0.0485, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 2.4601955688086847, | |
| "learning_rate": 2.0347474747474746e-07, | |
| "loss": 0.0437, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.9127272727272727, | |
| "grad_norm": 6.845607279107924, | |
| "learning_rate": 1.9539393939393938e-07, | |
| "loss": 0.0456, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 0.9163636363636364, | |
| "grad_norm": 3.267607748623562, | |
| "learning_rate": 1.873131313131313e-07, | |
| "loss": 0.0451, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 2.1934900287360666, | |
| "learning_rate": 1.7923232323232322e-07, | |
| "loss": 0.0461, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 0.9236363636363636, | |
| "grad_norm": 2.090160880581652, | |
| "learning_rate": 1.7115151515151512e-07, | |
| "loss": 0.0505, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 0.9272727272727272, | |
| "grad_norm": 4.966097333564902, | |
| "learning_rate": 1.6307070707070704e-07, | |
| "loss": 0.0537, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 0.9309090909090909, | |
| "grad_norm": 2.1214979148493076, | |
| "learning_rate": 1.5498989898989896e-07, | |
| "loss": 0.0461, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 0.9345454545454546, | |
| "grad_norm": 4.331745712735202, | |
| "learning_rate": 1.4690909090909089e-07, | |
| "loss": 0.0499, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 0.9381818181818182, | |
| "grad_norm": 8.581243563077976, | |
| "learning_rate": 1.3882828282828284e-07, | |
| "loss": 0.0493, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 0.9418181818181818, | |
| "grad_norm": 4.054442881590284, | |
| "learning_rate": 1.3074747474747476e-07, | |
| "loss": 0.0556, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 0.9454545454545454, | |
| "grad_norm": 5.0357651442227285, | |
| "learning_rate": 1.2266666666666665e-07, | |
| "loss": 0.0523, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.9490909090909091, | |
| "grad_norm": 4.173721876206746, | |
| "learning_rate": 1.1458585858585858e-07, | |
| "loss": 0.042, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 0.9527272727272728, | |
| "grad_norm": 5.514831485888669, | |
| "learning_rate": 1.065050505050505e-07, | |
| "loss": 0.0492, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 0.9563636363636364, | |
| "grad_norm": 5.89925297032685, | |
| "learning_rate": 9.842424242424242e-08, | |
| "loss": 0.0493, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 6.457997696881727, | |
| "learning_rate": 9.034343434343433e-08, | |
| "loss": 0.0516, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 0.9636363636363636, | |
| "grad_norm": 10.243171836638412, | |
| "learning_rate": 8.226262626262625e-08, | |
| "loss": 0.0529, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 0.9672727272727273, | |
| "grad_norm": 2.4837463351233695, | |
| "learning_rate": 7.418181818181817e-08, | |
| "loss": 0.0418, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 0.9709090909090909, | |
| "grad_norm": 7.115315330310798, | |
| "learning_rate": 6.61010101010101e-08, | |
| "loss": 0.0537, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 0.9745454545454545, | |
| "grad_norm": 7.137360943243018, | |
| "learning_rate": 5.818181818181818e-08, | |
| "loss": 0.0446, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 0.9781818181818182, | |
| "grad_norm": 5.279850881667171, | |
| "learning_rate": 5.0101010101010105e-08, | |
| "loss": 0.0504, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 0.9818181818181818, | |
| "grad_norm": 4.781435096498879, | |
| "learning_rate": 4.202020202020202e-08, | |
| "loss": 0.0519, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.9854545454545455, | |
| "grad_norm": 4.141078938486959, | |
| "learning_rate": 3.393939393939394e-08, | |
| "loss": 0.057, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 0.9890909090909091, | |
| "grad_norm": 3.983226535807257, | |
| "learning_rate": 2.5858585858585858e-08, | |
| "loss": 0.0588, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 0.9927272727272727, | |
| "grad_norm": 8.448308055260652, | |
| "learning_rate": 1.7777777777777777e-08, | |
| "loss": 0.0578, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 0.9963636363636363, | |
| "grad_norm": 2.0151546432182283, | |
| "learning_rate": 9.696969696969696e-09, | |
| "loss": 0.0503, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.6109640799264664, | |
| "learning_rate": 1.6161616161616161e-09, | |
| "loss": 0.0573, | |
| "step": 13750 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 13750, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 3000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |